rspde.matern {rSPDE} | R Documentation |
Matern rSPDE model object for INLA
Description
Creates an INLA object for a stationary Matern model with general smoothness parameter.
Usage
rspde.matern(
mesh,
nu.upper.bound = 4,
rspde.order = 2,
nu = NULL,
B.sigma = matrix(c(0, 1, 0), 1, 3),
B.range = matrix(c(0, 0, 1), 1, 3),
parameterization = c("spde", "matern", "matern2"),
B.tau = matrix(c(0, 1, 0), 1, 3),
B.kappa = matrix(c(0, 0, 1), 1, 3),
start.nu = NULL,
start.theta = NULL,
prior.nu = NULL,
theta.prior.mean = NULL,
theta.prior.prec = 0.1,
prior.std.dev.nominal = 1,
prior.range.nominal = NULL,
prior.kappa.mean = NULL,
prior.tau.mean = NULL,
start.lstd.dev = NULL,
start.lrange = NULL,
start.ltau = NULL,
start.lkappa = NULL,
prior.theta.param = c("theta", "spde"),
prior.nu.dist = c("beta", "lognormal"),
nu.prec.inc = 1,
type.rational.approx = c("chebfun", "brasil", "chebfunLB"),
debug = FALSE,
shared_lib = "detect",
...
)
Arguments
mesh |
The mesh to build the model. It can be an |
nu.upper.bound |
Upper bound for the smoothness parameter. |
rspde.order |
The order of the covariance-based rational SPDE approach. |
nu |
If nu is set to a parameter, nu will be kept fixed and will not
be estimated. If nu is |
B.sigma |
Matrix with specification of log-linear model for |
B.range |
Matrix with specification of log-linear model for |
parameterization |
Which parameterization to use? |
B.tau |
Matrix with specification of log-linear model for |
B.kappa |
Matrix with specification of log-linear model for |
start.nu |
Starting value for nu. |
start.theta |
Starting values for the model parameters. In the stationary case, if |
prior.nu |
a list containing the elements |
theta.prior.mean |
A vector for the mean priors of |
theta.prior.prec |
A precision matrix for the prior of |
prior.std.dev.nominal |
Prior std. deviation to be used for the priors and for the starting values. |
prior.range.nominal |
Prior range to be used for the priors and for the starting values. |
prior.kappa.mean |
Prior kappa to be used for the priors and for the starting values. |
prior.tau.mean |
Prior tau to be used for the priors and for the starting values. |
start.lstd.dev |
Starting value for log of std. deviation. Will not be used if start.ltau is non-null. Will be only used in the stationary case and if |
start.lrange |
Starting value for log of range. Will not be used if start.lkappa is non-null. Will be only used in the stationary case and if |
start.ltau |
Starting value for log of tau. Will be only used in the stationary case and if |
start.lkappa |
Starting value for log of kappa. Will be only used in the stationary case and if |
prior.theta.param |
Should the lognormal prior be on |
prior.nu.dist |
The distribution of the smoothness parameter. The current options are "beta" or "lognormal". The default is "lognormal". |
nu.prec.inc |
Amount to increase the precision in the beta prior distribution. Check details below. |
type.rational.approx |
Which type of rational approximation should be used? The current types are "chebfun", "brasil" or "chebfunLB". |
debug |
INLA debug argument |
shared_lib |
Which shared lib to use for the cgeneric implementation? If "detect", it will check if the shared lib exists locally, in which case it will use it. Otherwise it will use INLA's shared library. If "INLA", it will use the shared lib from INLA's installation. If 'rSPDE', then it will use the local installation (does not work if your installation is from CRAN). Otherwise, you can directly supply the path of the .so (or .dll) file. |
... |
Only being used internally. |
prior.kappa |
a |
prior.tau |
a list containing the elements |
prior.range |
a |
prior.std.dev |
a |
Value
An INLA model.